In witty, imaginative minds, on the other hand, the routine is broken through with ease at any moment; and one field of mental objects will suggest another with which perhaps in the whole history of human thinking it had never once before been coupled. The link here is usually some _analogy_ between the objects successively thought of,--an analogy often so subtle that, although we feel it, we can with difficulty analyze its ground; as where, for example, we find something masculine in the color red and something feminine in the color pale blue, or where, of three human beings' characters, one will remind us of a cat, another of a dog, the third perhaps of a cow. * * * * * Psychologists have of course gone very deeply into the question of what the causes of association may be; and some of them have tried to show that contiguity and similarity are not two radically diverse laws, but that either presupposes the presence of the other. I myself am disposed to think that the phenomena of association depend on our cerebral constitution, and are not immediate consequences of our being rational beings. In other words, when we shall have become disembodied spirits, it may be that our trains of consciousness will follow different laws. These questions are discussed in the books on psychology, and I hope that some of you will be interested in following them there. But I will, on the present occasion, ignore them entirely; for, as teachers, it is the _fact_ of association that practically concerns you, let its grounds be spiritual or cerebral, or what they may, and let its laws be reducible, or non-reducible, to one. Your pupils, whatever else they are, are at any rate little pieces of associating machinery. Their education consists in the organizing within them of determinate tendencies to associate one thing with another,--impressions with consequences, these with reactions, those with results, and so on indefinitely. The more copious the associative systems, the completer the individual's adaptations to the world. The teacher can formulate his function to himself therefore in terms of 'association' as well as in terms of 'native and acquired reaction.' It is mainly that of _building up useful systems of association_ in the pupil's mind. This description sounds wider than the one I began by giving. But, when one thinks that our trains of association, whatever they may be, normally issue in acquired reactions or behavior, one sees that in a general way the same mass of facts is covered by both formulas.